On November 17, 1999, at 1649 hours mountain standard time, a Hawker Siddeley Hunter T MK 7 jet, N576NL, collided with a ditch following a complete power loss on initial takeoff at Williams-Gateway Airport, Mesa, Arizona. The airplane, owned by Mach Two Inc. and operated by Advanced Training Systems International, Inc., sustained substantial damage. The commercial pilot sustained serious injuries and his passenger received minor injuries. The flight was being conducted under 14 CFR Part 91 when the accident occurred. The purpose of the flight was to reposition the aircraft to California for training purposes. Visual meteorological conditions prevailed at the time of the accident and the airplane was on an IFR flight plan. The flight originated at the airport moments before the accident. Use your browsers 'back' function to return to synopsisReturn to Query Page
The pilot said that the airplane was prepared for flight and serviced by his crew chief before he did a preflight on the airplane. He stated that everything appeared ok and the airplane appeared ready for flight. The pilot said the engine start was uneventful, "nice and cool," with temperatures staying below 500 degrees as the engine spooled up to idle power in less than 20 seconds. He requested and received an IFR clearance to NAS Pt. Mugu from ground control, went through the post start and taxi checks, and taxied out to runway 30C. After they reached the hold short line, they completed the takeoff checklist and called for takeoff. They were cleared for takeoff with a right turnout after departure.
The pilot said they taxied out on the runway and ran the engine up to 7,200 rpm and checked the instruments and flight controls. The engine temperature, oil pressure, hydraulic pressure, and boosted flight controls were working well. He then released the brakes, pushed the throttle up to the "military power detent," and checked the instruments again as they began the takeoff roll. He stated that rpm was 8,100, temperature was 650 degrees, and oil and hydraulic pressure was good. He said the airplane accelerated nicely and he rotated at 130 knots and flew off at 140 knots after using about 3,000 feet of take-off roll.
About 5-10 seconds after takeoff, the pilot said he experienced a significant compressor stall with the engine "chug" and a loud explosion. He notified the crew chief that they had just lost the engine, and that he was going to set the airplane back down. He lowered the landing gear and set up for a normal touchdown back on the runway. He extended the drag chute just before touchdown and said he felt a "good tug." Once back on the ground, he immediately began applying the brakes in an attempt to stop the airplane before they reached the flood control ditch at the end of the runway. He was unable to stop the airplane before it hit the ditch and they went into the ditch and back out the other side. When the airplane came to rest, the pilot said he told his crew chief that he had broken his back and asked him to blow the canopy for him to get them out of the airplane.
The crew chief, positioned in the right seat, stated that the preflight, start-up, run-up, and takeoff were "normal," and he did not realize that there was a problem until the pilot informed him that the engine had quit. He said he did not recall seeing any panel lights or hearing any aural warnings.
An eyewitness to the crash said he was located on the south ramp at the airport and that the ramp was approximately 45-degrees to the runway. His line-of-sight had him watching the middle of the takeoff run. He said he heard a "bang" which sounded like a compressor stall and then the airplane was out of his sight. He said he heard two more compressor stalls at close to the same magnitude. He said he would estimate the airplane was at 100 feet agl when he first noticed the takeoff.
A witness, who was a pilot in an airplane behind the accident airplane, was interviewed. He stated he observed a 30-foot flame exit the engine when it was about 200 feet above the ground on the initial takeoff climb. The pilot said that he did not observe any catastrophic failure of the engine or any parts depart the engine as he observed the flames.
One eyewitness who watched the airplane takeoff stated he heard a "bang" that sounded like a compressor stall after the airplane had climbed to approximately 100 feet agl. He said the airplane continued out of his sight but that he heard at least two other loud noises very similar to the first noise after the airplane disappeared from his sight.
The airplane is a two-seat ex-military jet fighter produced in the 1950's and early 1960's by Hawker for service in the United Kingdom Royal Air Force. Following the airplane's surplus from the RAF, it was imported into the United States and was being flown under a Federal Aviation Administration (FAA) special airworthiness certificate in the experimental category for the purpose of exhibition and racing. The airworthiness certificate was issued on December 12, 1994, by the Minneapolis, Minnesota, Flight Standards District Office.
The airplane's logbook shows three recent flights including the ferry flight from Minnesota to Williams Gateway on November 9, 1999, for a total of 3.5 hours. The pilot of the ferry flights was also the pilot on the accident flight.
According to Rolls-Royce, a review of their records indicated that the engine, serial number 5919, was manufactured in 1954 and sent to the RAF as part of a spares contract. The engine was last dispatched from their overhaul facility on March 26, 1981, following part life rework (overhaul) at 251 hours technical service order (TSO). According to Rolls-Royce, the engine is life limited (overhaul requirement) at 450 hours from parts rework and it would be life expired at 701 hours TSO. The last records on this engine documented its installation in another Hawker Hunter, serial number XL617, by the RAF on September 28, 1983, with 392 hours TSO.
The engine maintenance records presented by the operator at the request of Safety Board investigators began on June 27, 1990, and documented its installation in the accident airframe at Volk Field, Wisconsin. An unsigned handwritten note preceded this entry and stated that the engine had a total run time since overhaul of 436 hours, with 265 hours remaining to life limit. No documentation was available to support the engine history between leaving RAF service and the installation in the accident airframe. The last entry in the records was dated July 2, 1999, and consisted of an annual inspection at a time of 548 hours since overhaul.
Safety Board investigators and technical representatives of Rolls-Royce examined the airplane on January 11-12, 2000. The salvage operator who recovered the airplane reported that the main fuselage fuel tanks and 100-gallon tip tanks were full of fuel, with just in excess of 400 gallons pumped into barrels at the crash site. The wings were removed by the salvage operator to assist in the recovery of the airplane. The nose gear was collapsed and the nose bay was damaged. Impact damage was observed to the forward section of the fuselage. Inspection of the intake prior to the engine removal showed that the intake ducting contained a moderate amount of dirt, with heavier concentrations present in the right duct. The starter fairing was intact with no apparent damage. The left portion of the instrument panel just above the avionics rack was buckled. All circuit breakers were in. The LP and HP throttle cocks were in the aft (cutoff) position. The fuel isolation switch was guarded and in the normal position.
The engine remained contained within the engine bay during the impact sequence. Further examination of the engine revealed that the forward roller track was fractured. The bottom of the exhaust duct showed upward bending. The aft portion of the tail cone contained evidence of fire damage external to the tail cone. Material burn through was evident extending from an area at the 6 o'clock position to the 3 o'clock position as viewed from the rear. The tail cone also had impact deformation. Inspection of the exhaust duct area revealed a heavy deposit of black foreign material between blades of the LP turbine. The deposit was noted to be centrifuged to the shroud ID and was brittle in nature. Additionally, the exterior contained a black brittle appearance. The deposits broke into pieces when removed with a small screwdriver.
Control continuity was established to the LP and HP throttle levers at the FCU prior to engine removal.
The engine accessories were found undamaged after the engine was removed from the engine compartment bay. All lines were found secure with consistent evidence of safety wire usage. The electrical connection at the fuel pressure-warning switch was found secure. The Intake Guide Vane Ram Assembly (IGV) was secure. Inspection of the air inlet showed damage to the IGV plus first and second stage compressors. The IGV showed FOD damage to the vane leading edge as well as the vane chord, pressure side. Several of the first and second stage compressor blades showed FOD damage to the leading edges, predominately mid-length to tip. The engine would not turn by hand.
A partial engine disassembly was then conducted in an attempt to locate the source of the mechanical lockup. Three of the eight combustion chambers were removed and inspected. All of the chambers contained moderate concentrations of metallizing (splatter) affixed to the chamber walls adjacent to the fuel nozzle ejector exit. Small flakes of a silver material were noted lying loose in the chamber. After removal of the combustion chambers, the leading edges of the HP Turbine was noted to contain heavy concentrations of a brittle black deposit formed around the blades. With the combustion chambers removed, the compressor outlet from the 12th stage could be viewed. Several broken pieces of what appeared to be compressor vane or blade material could be seen lodged between the outlet guide vanes. The air intake casing was removed followed by the upper compressor casing. As the upper casing was lifted by crane from the lower casing, several compressor blades/vanes fell onto the ground.
Removal of the upper half of the compressor casing revealed extensive damage to the compressor. Impact damage was observed on the stage 1 to stage 4 blades. Three blades from stage 3 had fractured through the root-fixing lug. The bottom portions of the lugs were retained in position by the securing pins. The top half of the lugs and the aerofoil had been lost. All the stage 5 to 10 blades had fractured through the aerofoil just outboard of the root platforms. Two stage 11 blades and a stage 12 blade had also fractured through the aerofoil just outboard of the root platforms. The remainders of the stage 11 and 12 blades were intact but exhibited impact damage. Most of the broken blades had been retained within the compressor section.
TESTS AND RESEARCH
Arrangements were made to collect fractured blades and vanes for laboratory analysis. Additionally, the salvage operator cut out two portions of the third stage compressor blade lug fractures and samples of the third stage compressor blades. These samples were sent to the Air Accidents Investigation Branch (AAIB) in Farnborough, England, to be taken to the Rolls-Royce Bristol facility for analysis.
According to the Rolls-Royce engine failure investigation report dated June 6, 2000, the primary failure was most likely the three stage 3 compressor blades that had failed in the root fixing lugs. The inspection also revealed extensive secondary damage to the compressor. The second stage blades exhibited extensive damage that was either FOD or secondary in nature.
Examination of the fractured lug revealed that it had broken at the 3 and 9 o'clock position. According to the Rolls-Royce report, their experience shows these are generally the positions of peak steady stress. Examination of the fracture face revealed that it had failed due to high cycle fatigue development that had initiated in the bore of the pin-fixing hole. The fatigue had propagated roughly 90 percent through the lug at the 9 o'clock position prior to final "break off." There was no evidence of fatigue development in the lug at the 3 o'clock position. The rear face of the lug fragment between the 6 and 9 o'clock position exhibited extensive intergranular cracking and pitting. Analysis using an Energy Dispersive X-ray (EDAX) technique revealed high levels of oxygen associated with the cracks and pits indicative of corrosion. At the fatigue nucleus, the corrosion had undermined a number of surface grains and at least one underlying grain of material. The grains had fallen away from the lug leaving a cavity 0.0134-inch deep at the nucleus. The report opined that the fatigue had nucleated at the intergranular corrosion.
The side faces of the fixing lug displayed a 0.118-inch wide band of heavy fretting damage located at the bottom of the lug. Similar damage was observed on the fixing lugs of the unbroken blade. The report opined that the damage was caused by excessive vibration and probably occurred after the three blades had failed.
Examination of the other fractured fixing lug (blade 2) revealed that it had also failed at the 3 and 9 o'clock positions. Examination of the fracture face revealed that it had failed due to high cycle fatigue development that had initiated in the bore of the pin-fixing hole. The fatigue had propagated roughly 75 percent through the lug at the 9 o'clock position prior to final fracture. Secondary fatigue systems were observed at the 9 o'clock position in the region of final break off.
The history of the stage 3 compressor blade failures was examined. There have been three other previously investigated occurrences of stage 3 blade failures resulting from fractures through the root fixing lugs. All three failures had occurred due to high cycle fatigue that had initiated at corrosion pitting and had propagated through the root-fixing lug resulting in release of the aerofoil.
In addition, the laboratories at East Kilbride had examined five other occurrences of lug failures that had not been formally reported. The details of the occurrences are contained in the laboratory report.
The root-fixing lug from one of the unbroken stage three blades was sectioned, mounted, and lightly polished into the front face to facilitate metallographic examination. This revealed that the bore of the pinhole had pulled away from the top of the bush. Small areas of corrosion were observed on the front face of the lug, mainly located around the blend radius between the front face of the bore of the pin-fixing hole. The examination revealed a corrosion crack 0.004 inches deep, located at the 7 o'clock position. The material exhibited a satisfactory microstructure composed of evenly distributed insoluble iron/nickel precipitates in the aluminum matrix.
Hardness measurements conducted on an unbroken lug returned values between 147 and 151 Hv (10 kg), which satisfied the drawing requirement of 120 to 140 HB (132 to 153 HV 10kg). Hardness measurements on lug 1 also returned values between 145 and 150 Hv (10kg).
Rolls-Royce reported that in response to the history of failures of stage 3 compressor blades due to high cycle fatigue development from corrosion pits, two modifications were introduced. MOD5522 introduced a strengthened lug and MOD5528 introduced a silicon rubber sealant to the bush bore interface. From examination of the submitted components, the metallurgist concluded that these modifications had not been incorporated on this engine.
Rolls-Royce stated that the RAF had completed the introduction of MOD5522, and had probably completed the introduction of MOD5528 to their fleet by the end of 1986. The entry in the engine record sheets shows that the engine was in operation at RAF Brawdy in 1983. Rolls-Royce concluded that it might be assumed that engine 5919 left RAF service between 1983 and 1986 without incorporation of these modifications.